Dynamo-K: Kinematic Dynamo & Supersymmetry

• Dynamo-K is the kinematic dynamo framework that studies the linear induction of weak magnetic fields through eigenmode analysis in prescribed conducting flows.
• It reformulates magnetic amplification using differential forms and a stochastic evolution operator, revealing a correspondence between topological supersymmetry breaking and chaotic dynamics.
• Extensive numerical, laboratory, and stellar investigations validate Dynamo-K by quantifying thresholds, spectral behavior, and rotation–activity relations in various astrophysical settings.

Searching arXiv for the specified paper and closely related kinematic dynamo work to ground the article in the cited literature. arXiv search: "(Ovchinnikov et al., 2015) Kinematic Dynamo, Supersymmetry Breaking, and Chaos" Dynamo-K is naturally read as the kinematic dynamo (KD): the linear induction problem for a magnetic field in a prescribed conducting flow, posed in the regime where the magnetic field is too weak to modify the flow through the Lorentz force. In this setting one studies eigenmodes, growth rates, threshold behavior, and spectral structure rather than nonlinear saturation. In the formulation emphasized by the supersymmetric theory of stochastics, the KD operator is exactly a stochastic evolution operator on differential forms, so exponentially growing magnetic modes are reinterpreted as spontaneous breaking of topological supersymmetry. Direct numerical simulations, laboratory-oriented kinematic calculations, Babcock–Leighton models, and stellar rotation–activity analyses all elaborate different aspects of this same linear framework (Ovchinnikov et al., 2015, Subramanian et al., 2014, 0803.3261, Pinter et al., 2011, Vashishth et al., 2021, Houdebine et al., 2017).

1. Linear induction problem and kinematic regime

In its standard form, the kinematic dynamo evolves the magnetic field $\mathbf{B}$ according to

$$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$

with prescribed flow $\mathbf{v}$ and magnetic diffusivity $\eta = 1/(\sigma\mu)$. “Kinematic” means precisely that $\mathbf{v}$ is externally given and the problem is linear in $\mathbf{B}$; one therefore asks whether there exist modes $\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$ with $\mathrm{Re}\,\lambda>0$. The same distinction appears across several settings: in turbulent DNS the field grows or decays exponentially in a statistically stationary flow, while in Babcock–Leighton mean-field models the large-scale flows and transport coefficients are prescribed and nonlinearity enters only through prescribed quenching functions rather than Lorentz-force feedback (Ovchinnikov et al., 2015, Subramanian et al., 2014, Vashishth et al., 2021).

This linearity gives Dynamo-K a double character. On one hand it is a minimal model of astrophysical field amplification, because it isolates stretching, advection, and diffusion. On the other hand it is a spectral problem: the dominant long-time behavior is controlled by the leading eigenmode of the induction operator. In helical turbulence this can include both small-scale and large-scale signatures within a single global eigenfunction; in laboratory and stellar contexts it enables clean threshold calculations even when the full nonlinear system is inaccessible (Subramanian et al., 2014).

2. Differential forms and the KD–STS correspondence

A central reformulation replaces the vector field $\mathbf{B}$ by a magnetic 2-form

$$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$

with $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$0 for the magnetic vector potential 1-form $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$1. In this language the divergence-free constraint becomes the Bianchi identity $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$2, and the induction equation becomes

$$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$3

Using Cartan’s formula $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$4 and $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$5, the KD operator takes the graded-commutator form

$$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$6

This is the structural statement behind the topological supersymmetry of the kinematic dynamo (Ovchinnikov et al., 2015).

The same operator arises in the supersymmetric theory of stochastics from the stochastic differential equation

$$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$7

for isotropic additive Gaussian white noise. The stochastic evolution operator on differential forms is then exactly $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$8. In this correspondence, the induction equation for magnetic 2-forms is literally the evolution equation of the stochastic flow restricted to $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$9, while the exterior derivative $\mathbf{v}$0 plays the role of a BRST-like supercharge and satisfies $\mathbf{v}$1 (Ovchinnikov et al., 2015).

3. Spectra, supersymmetry breaking, chaos, and the absence of scalar dynamos

The spectral problem is posed as $\mathbf{v}$2. Time evolution is $\mathbf{v}$3, so $\mathbf{v}$4 corresponds to exponential growth and $\mathbf{v}$5 to oscillation. Because $\mathbf{v}$6, eigenstates organize into non-supersymmetric boson–fermion doublets and supersymmetric singlets representing nontrivial De Rham cohomology classes. The key structural result is that every supersymmetric state has zero eigenvalue, whereas any eigenstate with nonzero eigenvalue is non-supersymmetric. Consequently, whenever the fastest mode has $\mathbf{v}$7, the dynamical ground state is non-supersymmetric and topological supersymmetry is spontaneously broken (Ovchinnikov et al., 2015).

In Dynamo-K this gives an exact reformulation of magnetic amplification: genuine dynamo action is spontaneous supersymmetry breaking. A recurrent misconception is that chaos is required only for the ideal fast dynamo with $\mathbf{v}$8. The KD–STS correspondence extends the same necessity to the diffusive case: if the kinematic dynamo operates, then the underlying stochastic flow must lie in the supersymmetry-broken phase, which is the STS definition of chaos. The existence of both purely growing and growing-oscillatory kinematic modes further shows that the broken-symmetry ground state can have either real or complex eigenvalues (Ovchinnikov et al., 2015).

The same framework also yields a sharp restriction on what can and cannot dynamo. In three dimensions, 0-forms and 3-forms have supersymmetric ground states: the constant function in $\mathbf{v}$9 and the equilibrium probability density in $\eta = 1/(\sigma\mu)$0. These sectors therefore do not support dynamo-type exponential growth under stationary flows. Only the 1-form and 2-form sectors can host non-supersymmetric ground states with negative real eigenvalue. This is the formal content of the statement that there is no scalar dynamo, whereas magnetic fields and vector potentials can be amplified (Ovchinnikov et al., 2015).

4. Spectral structure in turbulent kinematic dynamos

DNS of helical turbulence in the kinematic stage show that the magnetic energy spectrum is shape-invariant to high precision:

$\eta = 1/(\sigma\mu)$1

with $\eta = 1/(\sigma\mu)$2 across the full resolved range from box scale to resistive scales. This means that the magnetic field behaves as a single global eigenfunction of the linear induction operator even when large-scale and small-scale dynamo signatures coexist. The spectrum still rises toward a peak near the resistive wavenumber $\eta = 1/(\sigma\mu)$3, so the rms field remains dominated by small scales during the kinematic stage (Subramanian et al., 2014).

Helicity does not remove that small-scale dominance. In helical runs, the decomposition

$\eta = 1/(\sigma\mu)$4

reveals a clean excess at the lowest wavenumbers in one polarized component, and suitably chosen planar averages isolate a growing mean field. Yet these large-scale signatures remain embedded in an eigenfunction whose energy peaks at resistive scales. Accordingly, the relative mean-field amplitude decreases with magnetic Reynolds number: for fully helical forcing at $\eta = 1/(\sigma\mu)$5, $\eta = 1/(\sigma\mu)$6 at intermediate $\eta = 1/(\sigma\mu)$7 and $\eta = 1/(\sigma\mu)$8 for larger $\eta = 1/(\sigma\mu)$9 (Subramanian et al., 2014).

The generalized Kazantsev analysis in the same work interprets this through two regimes. In Type I, the small-scale dynamo dominates and the large-scale field is a decaying tail of the same eigenfunction; in Type II, the large-scale $\mathbf{v}$0 mode dominates and small scales are produced by tangling of that field. The DNS at large $\mathbf{v}$1 are identified with Type I. A second important quantitative result is that with scale separation $\mathbf{v}$2, a non-helical small-scale dynamo is excited already at $\mathbf{v}$3 for $\mathbf{v}$4, markedly below earlier estimates obtained with smaller scale separation (Subramanian et al., 2014).

5. Numerical realizations, boundary conditions, and laboratory dynamos

A substantial branch of Dynamo-K is numerical and geometrical. In cylindrical domains, a hybrid finite-volume/boundary-element scheme solves the induction equation in the conducting interior while enforcing exact insulating exterior boundary conditions through a scalar-potential boundary integral formulation. The finite-volume part uses constrained transport on a staggered mesh, preserving $\mathbf{v}$5 to machine precision; the boundary-element part maps the normal magnetic field on the boundary to tangential components through a precomputed dense matrix. In the validation problem based on the Marié–Normand–Daviaud flow, this FV–BEM approach gives $\mathbf{v}$6, in good agreement with differential-equation and integral-equation approaches, whereas simplified vanishing-tangential-field conditions yield $\mathbf{v}$7 and therefore underestimate the threshold substantially (0803.3261).

The importance of flow topology and boundary treatment becomes even clearer in VKS-oriented kinematic simulations based on SPIV-measured mean von Kármán flows. The predicted kinematic mode is consistently a non-axisymmetric $\mathbf{v}$8 equatorial dipole, while the observed saturated VKS mode is mainly axisymmetric and axial. The discrepancy is treated as evidence that mean-flow kinematics alone are not the whole story; the papers explicitly point to non-axisymmetric turbulence, ferromagnetic impellers, and nonlinear Lorentz-force effects as missing ingredients (Pinter et al., 2011).

VKS configuration	$\mathbf{v}$9	Kinematic $\mathbf{B}$0
$\mathbf{B}$1 with annulus	0.8	35.6–37.7
$\mathbf{B}$2 without annulus	0.8	41.6–47.6
$\mathbf{B}$3 with annulus	0.62	$\mathbf{B}$4
$\mathbf{B}$5 without annulus	0.49	$\mathbf{B}$6
$\mathbf{B}$7 synthetic, no annulus	0.8	$\mathbf{B}$8

These values show that the poloidal-to-toroidal ratio $\mathbf{B}$9 and the presence of an annulus in the shear layer plane strongly control kinematic dynamo onset. In the favorable $\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$0 configurations, the simulated thresholds lie in qualitative agreement with the experiment, which shows dynamo action around $\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$1. In the unfavorable $\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$2 configurations, especially without annulus, the mean flow is too toroidal and no kinematic dynamo appears up to the largest simulated $\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$3. This makes Dynamo-K a sensitive diagnostic of flow geometry even when it does not reproduce the final saturated experimental mode (Pinter et al., 2011).

6. Subcritical and hysteretic kinematic dynamos in Babcock–Leighton models

In axisymmetric Babcock–Leighton flux-transport models, the large-scale magnetic field is decomposed into a poloidal potential $\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$4 and a toroidal field $\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$5, with prescribed differential rotation, meridional circulation, and radial diffusivity profile. The control parameter is the dynamo number

$\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$6

while the effective dynamo number becomes

$\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$7

Because the prescribed quenching functions make $\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$8 increase with $\mathbf{B}(\mathbf{x},t)\propto e^{\lambda t}$9 at small field and decrease as $\mathrm{Re}\,\lambda>0$0 at large field, it has a maximum at intermediate $\mathrm{Re}\,\lambda>0$1. This non-monotonicity is the mechanism behind subcriticality and hysteresis (Vashishth et al., 2021).

For weak seed fields, the local-quenching model becomes supercritical at $\mathrm{Re}\,\lambda>0$2, corresponding to $\mathrm{Re}\,\lambda>0$3. Below that threshold, infinitesimal fields decay. However, if one starts from a strong oscillatory state obtained just above threshold and continues downward in $\mathrm{Re}\,\lambda>0$4, a finite-amplitude oscillatory branch persists below $\mathrm{Re}\,\lambda>0$5: the model is subcritical. The same bistable structure appears in the nonlocal-quenching case, where both the Babcock–Leighton source and turbulent diffusivity are quenched by the toroidal field averaged over the base of the convection zone. In both variants the same control parameter admits either a decaying solution or a strong oscillatory solution depending on initial conditions (Vashishth et al., 2021).

Stochastic modulation of the Babcock–Leighton amplitude,

$\mathrm{Re}\,\lambda>0$6

with month-scale coherence time, then tests the robustness of that subcritical branch. In the nonlocal-quenching case, even fluctuation levels of order $\mathrm{Re}\,\lambda>0$7 can eventually destroy subcritical solutions; a representative run at $\mathrm{Re}\,\lambda>0$8, just below the nonlocal critical value $\mathrm{Re}\,\lambda>0$9, survives for about $\mathbf{B}$0 years with $\mathbf{B}$1 fluctuations and then collapses. By contrast, strongly supercritical runs remain active and can exhibit grand minima: at $\mathbf{B}$2 with $\mathbf{B}$3 fluctuations, a $\mathbf{B}$4-year integration contains about $\mathbf{B}$5 grand minima, while none are detected for $\mathbf{B}$6 (Vashishth et al., 2021).

7. Stellar rotation–activity constraints and late-K “Dynamo-K”

Rotation–activity correlations provide an observational constraint on stellar dynamo regimes from dK4 through dM4. For low-activity stars, the Ca II surface-flux RAC slopes cluster around a common value: $\mathbf{B}$7 at dK4, $\mathbf{B}$8 at dK6, $\mathbf{B}$9 at dM2, $$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$0 at dM3, and $$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$1 at dM4. For high-activity unsaturated stars the early-type slopes are much steeper—$$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$2 at dK4e, $$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$3 at dK6e, and $$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$4 at dM2e—but they collapse to approximately $$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$5 at dM3e and dM4e. The paper interprets this as evidence for different dynamo regimes across and around the transition to complete convection: interface-dynamo dominance in active late-K and early-M stars, and distributed-dynamo behavior becoming dominant by dM3–dM4 (Houdebine et al., 2017).

Late-K stars are especially informative. In $$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$6 versus Rossby number, dK4 stars follow the canonical F–G–K relation, whereas at a given Rossby number the mean $$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$7 values lie lower by a factor of $$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$8 in dK6, $$F = \frac{1}{2} F_{ij}\,dx^i \wedge dx^j, \qquad F_{ij} = \epsilon_{ijk} B^k,$$9 in dM2, $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$00 in dM3, and $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$01 in dM4. At the same time, the ratio of coronal to chromospheric heating, $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$02, increases by a factor of $$\partial_t \mathbf{B} = \nabla \times (\mathbf{v} \times \mathbf{B}) + \eta \nabla^2 \mathbf{B}, \qquad \nabla\cdot \mathbf{B} = 0,$$03 between dK4 and dM4. This suggests that the late-K regime is not merely a scaled solar dynamo: it is already the onset of the systematic decline in chromospheric efficiency and the redistribution of magnetic-energy dissipation that become pronounced in mid-M dwarfs (Houdebine et al., 2017).

In that sense, stellar “Dynamo-K” denotes a bridge regime. dK4 behaves as a solar-type extension of the standard rotation–activity relation, while dK6 already departs from it in normalization despite preserving similar rotation dependence. The observational picture is therefore consistent with a two-component interpretation: a distributed dynamo visible in the nearly universal low-activity RAC slope across dK4–dM4, and an interface-dynamo contribution that is prominent in active late-K and early-M stars but weakens as the radiative core disappears (Houdebine et al., 2017).